reporting from the cube or framework package

[nfldbunker]

hi, what do people suggest to report against the cube or framework modeled reportng data..What would be easier for the end user?

Is there any difference in level of reporting?

thanks

[david.stachon]

yes, there is quite a difference in the way the two behave.

...the type of reporting that your end users want to perform is ultimately the determining factor as to which you decide to deploy.

and, it doesn't have to be EITHER/OR ...you can deploy both.

[CognosPaul]

Generally if you know the fields you want to research, then a cube is better. But if you want to look up distinct and unknown values, a relational source is better.

For example, lets say you want to know how many customers bought TrialChef Water Bags in China in the years 2004 through 2007. A relational source could give you the requested information, but it would be easier and faster with a cube.

If, however, you want to list all of the customers names and addresses, you would need to use a relational source.

[redmist]

I have found that if there is a requirement of generting financial reports (MTD, YTD etc) cubes makes much more sense. making those kind of reports against a cube using MDX functions is very simple

[Gollapudi]

I will totally agree with you redmist, coz we are working with DatetoDate kind of (MTD, YTD) reports, for which Cubes are very good, for detailed reports I feel relational model is farbetter..